Refik Anadol opened Dataland in Los Angeles, marking the world’s first museum entirely dedicated to artificial intelligence arts, according to studio announcements. Located in a space featuring continuous LED surfaces spanning floors, walls, and ceilings, the museum replaces traditional screens with an immersive environment driven by biometric sensors. A wrist sensor developed through medical research measures a visitor’s heart rate, skin temperature, and emotional arousal, translating physiological data into shifting patterns of light, color, and movement across the room.
Generative Realities and Biosensor Integration
Visitors walking through Dataland encounter a generative reality where the exhibition responds directly to their physical presence, according to studio documentation. The sensory experience includes a scent diffusing system that combines 400,000 molecules through an AI model to generate shifting aromas, such as rain hitting dry earth. Sound design featuring wind patterns, rainfall, and more than 100,000 bird vocalizations creates an immersive acoustic environment. In its opening week, the museum recorded several thousand visitors, positioning itself as an interactive laboratory rather than a static gallery.
The Large Nature Model and Indigenous Collaboration
The core engine behind Dataland’s exhibitions is the Large Nature Model (LNM), a self-developed artificial intelligence system trained on ecological data rather than human text repositories, according to project disclosures. The conceptual origin of the project traces back to a collaboration with the indigenous Yawanawá community in the Brazilian Amazon rainforest, where Anadol and his partner spent several weeks in 2021. That research led to the 2023 NFT collection “Winds of Yawanawá,” which raised 2,5 Mio.
To build the current LNM, Anadol partnered with the Smithsonian Institution, the Natural History Museum in London, and the Cornell Lab of Ornithology, alongside data collection expeditions across sixteen rainforest regions supported by Google. The resulting open-source model processes petabytes of raw data regarding wind patterns, tree water levels, and bird calls to generate abstract visual compositions in real time.
Market Positioning and Technological Infrastructure
Dataland enters the immersive art exhibition sector, a market valued at 2,8 Mrd. US-$ by 2034. While existing market leaders like Japan’s teamLab or Paris’s Atelier des Lumières offer immersive digital displays, Anadol’s studio utilizes a proprietary, biosensorically controlled generative model. The studio’s operations are supported by technology partnerships with Google, Nvidia, Microsoft, Apple, Intel, IBM, and Siemens, alongside academic collaborations with MIT, UCLA, Harvard, and Stanford.
Addressing environmental concerns regarding machine learning operations, studio representatives note that the Large Nature Model runs on a Google Cloud server in Oregon powered by 87 percent renewable energy, with a typical visitor experience consuming roughly the electrical equivalent of charging a smartphone.
Keep reading